Wednesday, August 29, 2007

My Case Against Mortgage Lending

With the price of many mortgage lenders hitting record lows, many people are wondering if now is the time to get into this sector. I am going to spell out my reasoning against investing in this area. To clarify, I am talking about the mortgage lenders which skipped traditional banking and instead focused on originating and selling large volumes of loans.

The basis of my argument revolves around the idea that the "value-added" from this industry is minimal at best. I summarize the mortgage lending business model as this: borrow at market rates from banks, hire a staff dedicated to finding and originating loans, and then sell them to other investors for a profit, and repeat. The whole process uses heavy leverage to make the returns worthwhile. Basically, they're aim is to add value in the originating process by making more profitable loans, whether through higher interest and/or more security (For sub-prime and Alt-A, it was higher rates while hopefully maintaining security) The mortgage lenders could have gone three routes:

A) originate loans to good borrowers that banks would also lend to.
B) Loan to good borrowers that banks would not lend to.
C) originate loans to bad borrowers.

In C there is no viable long-term business model. In Situation A, the "value-added" is minimal- you save the borrower the time from visiting their local bank, and in return add another layer of frictional costs. On average, the mortgage lenders total cost to originate are about 2% of loan value, so you can make a rough estimate of $4,000 per loan in additional costs.

Only in B does there exist a niche market which appears worthwhile. But the market will always occupy only a small niche outside bank lending criteria. If the non-conforming business grew large and the loans being made were truly good loans, then banks could simply loosen their criteria and add new competition, forcing it back to a situation A.

So when the market for Sub-prime and Alt-A grew to 40% of loans originated in 2006, either things got out of hand, or banks fell behind a new shift in acceptable lending. I choose to believe in the former.

Monday, August 27, 2007

Chou Funds Semi-Annual Report

Francis Chou's 2007 Semi-Annual Report is out along with some interesting commentary. He also made some purchases during the semester, which as far as I could tell were mostly:
(Numbers based on $ cost)

NEW* Alpha Natural Resources 6.3 mil
NEW* Sprint-Nextel 9.4 mil
NEW* Media General 4.3 mil
Sun-Times Media Group Increased from 4.2 mil to 13.9 mil
Watson Pharmaceuticals Increased from 1.4 mil to 11.9 mil
NEW* IDT Corp. 1 mil
NEW* Primus Telecom Debt 25 mil

For reference, Chou manages a total portfolio of 420 million. Overstock continues to be one of his significant holdings, with 36 million invested into it. The new position in Primus debt is also a very large position by Chou's standards.

Saturday, August 25, 2007

Delving into Investing Theory

As I was thinking about the recent credit problems, I couldn't help but think there was a lesson to be learned for investing as a whole. Looking from today, we can say that some lending got completely out of hand. Since lending really is a form of investment, we can look at their rationale for making these loans to find the fault. Loan underwriting relaxed to a point where banks became less concerned with borrower's ability to pay, and instead focused on the asset price securing the loan. This can be seen in the rise in lending debt-income ratios, the percentage of income of the borrower going to pay off debts. Banks conventionally have a limit at 36%. Subprime lender Delta Financial, considered one of the better underwriters, went up to 55%; Other subprime lenders rarely disclosed their maximum limit. These lenders tried to make up for this by having lower Loan-to-Value ratios(LTV), which is the loan amount compared to the value of the house. This added more principal security.

The problem as I see it is one of financial misunderstanding. To me, an investment has an asset value and an asset cost. The asset value is simply the net present value of future cash flow, while the asset cost was what was paid to acquire the assets. So, a loan's "asset value" is really the discounted sum of the borrower's interest and principal payments, along with the cash-out value if or when the borower defaults. The asset cost though is simply the amount of money that was loaned. From this perspective, we can see that the problem was that lenders lost track of asset value when they became less concerned with the borrower's income and relied more on the underlying home price. If the homes could be sold for their full value this wouldn't be a problem, but there are very significant costs associated with foreclosures. So many mortgage companies made loans that were carried at face value when their real value was considerably less, given the likelihood of default.

Similarly, this same behavior can be seen during stock market bubbles. A stock is also worth the net present value of its future cashflow. This means the sum of the cash flow from the operating business as well as a final gain (or loss) if the company needs to be liquidated. In stock market bubbles, investors become less focused on the present value of the investments they purchase; instead, they are buying an asset at a certain cost and hoping it will appreciate when they sell it to someone else. The business' ability to justify the price through its cashflow becomes an afterthought. So for a long time, equity investments can be carried on the books at a price that is much higher than logic would justify.

Your goal in investing is to be buying when asset cost is significantly below asset value.
The beauty of Ben Graham's philosophy was its simplicity. His strategy was to invest in companies at two-thirds of their Net Current Asset Value (NCAV), or their current assets minus total liabilities. Current assets are expected to be converted into cash within 18 months, and they can usually be taken at near face value. Such an investment situation means that if the company stopped operations immediately, it would have enough cash to pay all of its obligations and then give what is left over to shareholder's for a decent return in a relatively short period of time- hence, you were buying a company for less than its NPV under a scenario where the fixed assets and the company were treated as worthless. So, there was a high level of cash security for this type of investment, yet the upside was also pretty limited. But Ben Graham knew that fixed asset values can be much different than their costs, and wanted to make a strategy that any Average Joe could use.

The days of finding pure NCAV ideas are getting slim these days as information has become much more available. Today, a investor must become much more accurate and comfortable in make income projections for a business. This means analyzing competitive advantages will play a much more important part in security analysis, and there will be the possibility for larger gains- and mistakes- for investors.

Monday, August 20, 2007

Flight to Safety

Dazel from the Berkshire Hathaway Shareholders board pointed out something that hasn't recieved much press- the recent dramatic fall in treasury yields, especially short-term. See this link.

Meanwhile in the world of computer-model trading,

"Wednesday is the type of day people will remember in quant-land for a very long time," said Mr. Rothman, a University of Chicago Ph.D. who ran a quantitative fund before joining Lehman Brothers. "Events that models only predicted would happen once in 10,000 years happened every day for three days."

More "black swans"!

Friday, August 17, 2007

Pope and Talbot Announces Wood Pulp Increase

"Pope & Talbot, Inc. (NYSE:POP - News) today announces a $20 price increase to its customers in North American and a $30 price increase to its customers in Europe until further notice.

The new prices will be $850 in the North American market and $830 in Europe for its northern bleached softwood (NBSK) grades."


This is continued good news for SFK and the pulp industry. Meanwhile, I count three NBSK pulp producers still in troubled waters, which i define as negative EBITDA's and high leverage. These are Pope and Talbot(825,000 tonnes/year), Tembec(800,000 tonnes), and Catalyst Paper(525,000 tonnes).

Update: Also today, Catalyst announced the shutdown of 320,000 tonnes of yearly pulp production due to the B.C. union strike, which is causing a limitation of fiber in Western Canada.

Thursday, August 16, 2007

Economic Recap

Back in March, I said the following:

"One can't underestimate the effects this could have. If many of these subprime loans prove unsustainable without the hope of refinancing, this could increase defaults, which could decrease home prices, which could spread the default risk up the credit quality ladder. Meanwhile, mortgage insurers will be affected, along with banks, which pretty much spreads out to everywhere. Consumer demand, which has been so reliant on asset monetization, can drop. That branches out to affect the whole economy."

Well, so far we are beginning to see the beginning of the effects on banks and mortgage insurers. (I should of included hedge funds). But it's important to remember that subprime is just one example of the overall credit bubble. If I had to make a case for overvaluation in the markets, I would focus on three things:

1. Historical P/E's

2. Historical Return on Equity
Average earnings for the the Dow Jones Industrial Average are 11% of the company's book value in any 20-year period between 1920 and 1986 (1920-39, 1921-40, 1922-41, etc.). "Average earnings as a function of book value barely varies in the slightest, and has remained basically immune to inflation, wars, massive changes in the tax code or any other external factor." Warren Buffett said something similar, although I think he said 12%. Over the last decade, the DJIA Return on Equity has averaged 18%, with it currently running at 23%.

3. Leverage in the Economy

When things turn the other way, debt can get very messy. As the new Economist says, "because this crisis taps so deeply into the newly devised structures of finance, anyone who says the worst is definitely over is either a fool or someone with a position to protect. As risk has become bewilderingly dispersed, so too has information. ...Nobody knows how messy the inevitable bankruptcies will turn out to be. What markets need now is time to piece that information back together. Time before the next wave strikes." Similarly, many people have been claiming that high quality names have gone on sale during this recent market drop. I think it is way too early to tell.

"I place economy among the first and most important of republican virtues, and debt as the greatest of the dangers to be feared." -Thomas Jefferson, 1816


Monday, August 13, 2007

ICICI Lombard

ICICI Lombard recently came out with its annual report for the 06-07 year. The numbers show a company that continues to be a great growth story- here are some of the numbers:

Financial Year2006-072005-06
Figures in nos.
No. of policies sold


No. of claims handled642,777
No. of employees4,7702,283
No. of offices220154

Financial Year 2004-05 2003-04 2002-03 2001-02
Figures in nos.
No. of policies sold 607,926 249,531 98,293 9,148
No. of claims handled 84,970 23,487 8,022 420
No. of employees 1,249 561 284 116
No. of offices 96 63 35 11

Fairfax currently has a 24% stake in the company and has it recorded on the balance sheet at a conservative valuation. I see the largest private insurer in a fast growing industry, and the chance to take a lot of business from inefficient government competition. This could be a much more significant part of Fairfax in the future.

Saturday, August 11, 2007

The Brick 2nd Quarter Results

EBITDA down 3.5% for the quarter, up 1.8% YTD.
4.8% Same-store sales growth
Total sales up 7.3%
3 more franchises opened during the quarter, bringing the total to 29.
“With seasonally higher sales in the second half of the year and continued focus on cost management, we believe that we are very well positioned to drive increased profits for the Brick Group.”


Thursday, August 09, 2007

A Closer Look at Exchange Bank

I wanted to show you what really caught my eye with Exchange Bank of Santa Rosa. And, I also wanted to practice using Excel. Below is the resulting chart I made:

Note first that some fields are obviously missing because I was unable to ge that data. Also, the Net loan losses for 2Q07 is an annualized number. Delinquent refers to 30+ days late loans, while non-performing is 90+ days late loans plus non accrual. Also note my excellent Excel skills. A few things stand out about these numbers.

One, their overall loan underwriting is superb. As a comparison, a look at Countrywide's servicing portfolio in the 2nd quarter showed 5.02% delinquencies, and 1.74% non-performing, and their portfolio is heavily adjustable. Second, they are also simultaneously very conservatively reserved. Most other banks I have looked up have a ratio of allowance to non-performing of 100 to 150 percent. Exchange Bank hasn't been under 300% for ages. (Countrywide is under 50%)

So despite my belief in a credit bubble, I am willing to invest in this bank because of their allowance cushion and their strict underwriting. This goes along with the other many qualitative aspects: Their number 1 county market share, low cost deposits, and a Return on Assets averaging 1.5% in an industry where 1% is considered good. (Countrywide is at 1.06%, and theyre also origination 460 billion in loans each year on only 14 billion in equity. I didn't intend to pick on Countrywide, but its just too easy. Fairfax does also own CDS's against their debt.) And to top it off, it is at 10x earnings. Now, due to the company's structure it can not get bought out, and there is a chance it might ride the financial momentum downward in price, but it would be an opportunity to add to a longer term holding.

Wednesday, August 08, 2007

SFK Pulp 2nd Quarter Conference Call and Recap

A lot of people are dissapointed with SFK's 2nd quarter report, so I thought I'd share my thoughts.
The two main things I think need to be addressed are the drop in sales volume, and the strengthening Canadian dollar and its effects.

Sales Volume
SFK's NBSK pulp sales volume was a very low 75,514 tonnes this quarter. In the conference call, the company addressed this due to a major customer cutting back purchases for the quarter. Someone on the call asked whether the drop was due to them being unable to shift the business fast enough, and management said it was partly that and partly a decision to wait until the customer came back, because the customer was close by and the transportation savings accrete to SFK. Regardless, I wouldn't be concerned because in the end pulp is still a commodity product, and one that is currently at low industrywide inventories. In the end, sales volume will be close to production, and in the call management said sales volume and inventories have already returned to normal. The NBSK Mill has a yearly production of 375,000 tonnes, so "normal" is about 93,750 tonnes per quarter. Note also that 2nd quarter and 4th quarter take maintenance downtime, so these quarters have lower production, offset by higher 1st and 3rd quarter production.

For those that want to know the effect, here's some basic math:
NBSK business cost of sales as total sales was 78.8% (47,628/ 60,414)
18,236 more tonnes x $800 CAN price per tonne = $14,589,000
21.2% x 14,589,000 = $3,093,000 extra free cashflow

Note also that this doesn't take into account the fact that labor and maintenance would not take additional expenses for the added sales volume, meaning our gross margin used is likely understating the extra free cash flow. But at the risk of becoming short-term Wall Street analyst-like, im going to not bother doing that calculation.

Strengthening of the Loony
This is where things that getting (more) complicated. The Canadian dollar strengthened from an average of .8535 to .9107 US/CAN. This has a few effects. First, since pulp prices are derived in US dollars, a stronger Canadian dollar means lower realized prices for SFK in its own currency, while its costs remain the same. But also, the company had an additional 3.8 million impairment of cash and accounts recievable that are denominated in U.S. dollars.

Also, and this seems to have been missed by many, is the effect of this on the RBK business. The RBK business is cost, revenues, and profit are all in US dollars, but since SFK reports in Canadian dollars, the income from this segment drops when the loony rises (in canadian dollar terms). But, to hedge this currency risk, SFK took the debt for this acquisition out in US dollars, so its debt obligations also fall as the Canadian dollar rises. However, due to accounting treatment, the company does not report this change on the income statement until the RBK business pays its first dividend to the parent company. So instead, SFK recorded a 5,451 million currency translation adjustment on its balance sheet to take into account the lower debt outstanding in Canadian dollar terms, but this never made it to the income statement (Unlike the 3.8 million impairment mentioned above)

Where Now?
The Canadian dollar has continued to increase since last quarter. Assuming the rate averages out to a new .95, we get the following "run rate" quarterly calculations:
1. Exchange rate at .95 US/CAN

2.$786 CAN NBSK revenue per tonne
2nd Quarter sales price was $800/tonne CAN. If you take into account an $18US rise in prices and the new exchange rate, the sales price for 3Q will be about $786.

3.$618 CAN NBSK cost per tonne
Cost of sales in 3Q is usually lower due to a lack of maintenance expense- it was $571 CAN 3Q06. But since this is run rate calculations, I will start from the 2006 cost of sales of $603 CAN. Costs are up 1.8% this year so far, but I added some leeway, giving a cost of sales of $618.

4. 93,750 tonnes sold per quarter
Again, due to the sales slowdown in 2Q, sales volume will likely be much higher than average next quarter, but since this is run rate im using the average number.

Run Rate
NBSK EBITDA 15,624,000
RBK EBITDA.. 3,500,000
SG&A Expense -1,350,000
Interest Expe. -4,000,000
Capital Expen. -4,000,000

Yields Free Cash Flow of $9,764,000. For those that were concerned, yes this run rate would mean a distribution cut seems likely. Oh, and also, the stronger Canadian dollar would also reduce debt obligation by 4,850,000 in the 3rd quarter, meaning when that currency adjustment is realized, that will be upward of 10 million now on the income statement, though that should also be ignored as far as income goes.

Those are the numbers. Now here's why I'm holding, and you can choose to agree or disagree from here. The company is still cheap, at about 10x cashflow under current conditions. But on a broader worldwide scale, the Eastern Canadian production is ripe for change. Canadian pulp does not need to be automatically considered at a disadvantage to worldwide production. Labor makes up a small percentage of costs, and in fact operational costs at most eastern Canadian mills are actually lower than other places due to very low energy cost. The main burden is fiber. Fiber makes up about $300 of the cost per tonne of pulp in Eastern Canada, about $125 in Western Canada, and significantly lower in Latin America. Canadian fiber doesnt need to be so expensive- rather, it is a problem of too much fiber demand in one region, escalating prices. The industry figures are below:
49 million tonnes worldwide hardwood pulp demand per year
11 million tonnes of pulp(all grades) production in Canada
12 million tonnes of pulp production in Europe including Nordic countries and Russia
9 million tonnes of pulp production in South America

13 million tonnes worldwide NBSK pulp demand per year
7 million tonnes NBSK pulp production in Canada

NBSK pulp is a particular grade that requires stronger fibers found in only some places (mostly, Canada). NBSK also accounts for 28.8% of pulp demand. As a grade, it will likely survive, and give Canadian producers a competitive advantage. But on a broader scale, the fiber disadvantage of Eastern Canada is taking its toll, and this is the first place that excess capacity is coming off. Looking at it today, both Tembec and Pope and Talbot seem discounted already for bankruptcy in the markets. Tembec has 825,000 NBSK production, 1,035,000 hardwood production; Pope & Talbot has 820,000 NBSK production. Also, the merger of Abitibi and Bowater combine two powerhouses in Eastern Canada, possibly spurring much needed rationalization. As a very low cost producer, obviously any fall in fiber prices will greatly benefit SFK. This may sound like wishful thinking to some, and you are slightly right- more research is needed to make a compelling case one way or the next, and hopefully a few calls can help answer some of my new questions. Yet, this year western Canadian fiber is up 57%, compared to only 3% in eastern Canada. A cheap valuation with a possible catalyst, good management, and secured by higher replacement value keeps me invested and feeling relatively safe.

Monday, August 06, 2007

A Peak Inside Fairfax's CDS Portfolio

Given the enormous volatility going on in Fairfax's CDS portfolio, I thought i might try to shed some light on the matter. Below are some of the major holdings for a subsidiary, Odyssey Re America, at end of December 2006. Prem has stated that some of these are looking for capital appreciation, while others are to hedge their reinsurance recoverables and business. The ones in bold are the ones I assume were for investment rather than hedging purposes.

Company / notional amount in millions
Ace Holdings - 110
Allianz France - 130
Societe Generale- 175
Aegon - 105
Swiss Re- 75
Ambac - 110
AIG (?) - 330
Countrywide- 130
Freddie Mac- 235
Hanover - 155
JP Morgan - 75
MBIA - 65
MGIC - 205
PMI Group - 230
Radian - 260
Washington Mutual - 210
XL Capital - 200

Most of the counterparties for these deals are Citibank, Duetsche Bank, and Barclays. I'm particularly glad to see Countrywide, MGIC, PMI, and Radian among the bigger holdings on that list, especially since I'm not too optimistic on their futures and because their debt protection costs have been soaring the past few days. I'll be watching these names closely for updates.

Sunday, August 05, 2007

2nd Quarter Earnings Update

Earnings season has come and mostly gone for the companies in my portfolio. Here's a brief overview.

The Bad
The company reported 1 million in net earnings, or $.20 diluted EPS.
Shareholder's equity increased to 37.41 million, or $7.406 per share.
The big negative for this quarter was the .5 million increase in reserves for the discontinued bond program, as well as the addition of this statement:

"Highlands has provided claim information to the Company with respect to alleged losses during 2001 and 2002 for bail bonds issued in the State of New Jersey and for federal immigration bonds. Highlands has indicated in filings that it has additional exposure for bail bonds issued in states other than New Jersey. Highlands has not provided sufficient information for the Company to quantify certain of these additional losses or allocate such losses among the 2001 and 2002 years in which the Company participated and the 2000 year in which the Company did not participate. As of June 30, 2007, the Company is reserving to its best estimate of future Highlands losses based on the most recent loss information received from Highlands with respect to immigration bonds and New Jersey bail bonds only."

Also, commission expense continues to go up as business shifts to other products.
The one positive is that the company started providing more information about their premiums, including how much is ceded to the reinsurance companies. For the six months ended june 30, written premiums ceded were 14.29 million, compared to 4.16 million last year.

SFK Pulp Fund
SFK earned distributable cash of only 7.6 million for the quarter, although several factors were affecting the figure.

On the NBSK side- besides it being a maintenance shutdown quarter, the company also lost some business from some customers and was unable to replace it in time. Sales volume in the quarter was only 75,514 tonnes, while production per quarter for the mill averages approximately 93,000. Since NBSK pulp is a commodity product currently at record low inventory levels (24 days according to Canfor), I am not worried about the quarterly drop in volume.

The other main factor was the continued strengthening of the Canadian dollar. Quarter over Quarter prices actually fell from $821 CAN to $800CAN, even though prices have been rising in US terms. Also, the company took an additional 3.8 million charge from writing down U.S. accounts recievables.

The main thesis behind SFK still remains intact. Although the Canadian dollar has continued to increase since the 2nd quarter, this affects all other Canadian pulp mill operators equally, and Econ 101 tells us producers will continue to curtail production. (The most recent announcement being from Pope and Talbot for 68,000 tonnes NBSK pulp) Also, we are seeing the fiber price imbalance between Western and Eastern Canada also starting to reverse, which will put additional upward pressure on pulp prices. There will be a profitable spot for a low cost producer like SFK in such an environment when all is said and done.

The Good
Fairfax Financial
FFH has come along way and things are really starting to shine.
Underwriting income for the quarter increased to 87.2 million, net earnings per share of $8.92.
Book value increased to $165.50 per share, debt continues to be paid off or extended.
I've long stated my agreement with Prem's prediction of a one in fifty year event coming along in the markets. The positioning of his portfolio looks brilliant right now.
-Majority in cash and long term US government bonds, and no exposure to mortgage securities.
-Equity portfolio is 80% hedged with shorts against the market or specific stocks.
-a large CDS portfolio against financial services companies. At the end of the quarter this had a market value of 198 million. By the end of July, given recent market events, the market value of this has increased to 537 million, and so far since August the move has been significant as well.
Also, so far the hurricane season has been benign, and if things continue this way we can look for a positive reserve adjustment by year end.

Exchange Bank of Santa Rosa
YTD net income of $6.59 per share, Book value of $78.16 per share.
Continued excellent loan performance and conservative accounting:
-Nonperforming loans as % of total loans at .41%
-Total Delinquent loans as % of total loans at .66%
-Allowance for loan losses of 1.81%

As I am writing this, I'm realizing that I have never done a complete and thorough write-up on Exchange Bank. Many of you are probably wondering why I would invest in a bank given my negative thoughts on the financial industry, so look forward for some clarification soon.